Merge branch 'release-2.0.0-alpha'

.ci/daint.cscs.ch/Jenkinsfile

@@ -0,0 +1,42 @@
+pipeline {
+    agent any
+
+    stages {
+        stage('prepare') {
+            steps {
+                checkout scm
+            }
+        }
+        stage('build') {
+            steps {
+                run_batch("0:15:00")
+            }
+        }
+        stage('test') {
+            steps {
+                run_batch("1:00:00")
+            }
+        }
+    }
+}
+
+def run_batch(timelimit) {
+    def (account, basename) = env.JOB_NAME.split('/')
+    // avoid using the shell for variable expansion to
+    // get the final command displayed in Jenkins
+    try {
+        sh """
+        sbatch --wait \
+            --time="${timelimit}" \
+            --account="${account}" \
+            --job-name="${basename}" \
+            --output="sbatch.${env.BUILD_TAG}.${env.STAGE_NAME}.out" \
+            .ci/daint.cscs.ch/${env.STAGE_NAME}.sh
+        """
+    }
+    finally {
+        echo readFile("sbatch.${env.BUILD_TAG}.${env.STAGE_NAME}.out")
+    }
+}
+
+// vim: set filetype=groovy ts=4 sw=4 tw=0 :

.ci/daint.cscs.ch/build.sh

@@ -0,0 +1,36 @@
+#!/bin/bash -l
+
+#SBATCH --export=ALL
+#SBATCH --exclusive
+#SBATCH --constraint="mc"
+#SBATCH --partition="cscsci"
+#SBATCH --nodes=1
+#SBATCH --ntasks-per-node=4
+#SBATCH --cpus-per-task=3
+#SBATCH --ntasks-per-core=1 # 1=no HT, 2=HT
+
+set -o errexit
+set -o nounset
+set -o pipefail
+
+module swap PrgEnv-cray PrgEnv-gnu
+module load daint-gpu cudatoolkit CMake/3.12.0
+module unload cray-libsci_acc
+
+set -o xtrace  # do not set earlier to avoid noise from module
+
+umask 0002  # make sure group members can access the data
+
+mkdir --mode=0775 -p "${SCRATCH}/${BUILD_TAG}"
+cd "${SCRATCH}/${BUILD_TAG}"
+
+cmake \
+    -DUSE_CUDA=ON \
+    -DUSE_CUBLAS=ON \
+    -DWITH_GPU=P100 \
+    -DMPIEXEC_EXECUTABLE="$(command -v srun)" \
+    -DTEST_MPI_RANKS=${SLURM_NTASKS} \
+    -DTEST_OMP_THREADS=${SLURM_CPUS_PER_TASK} \
+    "${WORKSPACE}" |& tee -a "${STAGE_NAME}.out"
+
+make VERBOSE=1 -j |& tee -a "${STAGE_NAME}.out"

.ci/daint.cscs.ch/test.sh

@@ -0,0 +1,34 @@
+#!/bin/bash -l
+
+#SBATCH --export=ALL
+#SBATCH --exclusive
+#SBATCH --constraint="gpu"
+#SBATCH --partition="cscsci"
+#SBATCH --nodes=1
+#SBATCH --ntasks-per-node=4
+#SBATCH --cpus-per-task=3
+#SBATCH --ntasks-per-core=1 # 1=no HT, 2=HT
+
+set -o errexit
+set -o nounset
+set -o pipefail
+
+module swap PrgEnv-cray PrgEnv-gnu
+module load daint-gpu cudatoolkit CMake/3.12.0
+module unload cray-libsci_acc
+
+set -o xtrace  # do not set earlier to avoid noise from module
+
+umask 0002  # make sure group members can access the data
+
+mkdir --mode=0775 -p "${SCRATCH}/${BUILD_TAG}"
+cd "${SCRATCH}/${BUILD_TAG}"
+
+export CRAY_CUDA_MPS=1 # enable the CUDA proxy for MPI+CUDA
+export OMP_PROC_BIND=TRUE # set thread affinity
+# OMP_NUM_THREADS is set by cmake
+
+# document the current environment
+env |& tee -a "${STAGE_NAME}.out"
+
+env CTEST_OUTPUT_ON_FAILURE=1 make test |& tee -a "${STAGE_NAME}.out"

.flake8

@@ -1,2 +1,4 @@
 [flake8]
-max-line-length = 120
+max-line-length = 128
+# ignore whitespaces in array slices (https://github.com/PyCQA/pycodestyle/issues/373)
+ignore = E203,W503

.pre-commit-config.yaml

@@ -4,23 +4,20 @@ default_language_version:
 exclude: '^tools/(prettify/fprettify|build_utils/fypp)'
 fail_fast: false
 repos:
-- repo: https://github.com/pre-commit/mirrors-yapf
-  rev: v0.24.0
+- repo: https://github.com/ambv/black
+  rev: stable
   hooks:
-  - id: yapf
+  - id: black
     language_version: "python3"
 - repo: https://github.com/pre-commit/pre-commit-hooks
   rev: v2.0.0
   hooks:
-  - id: flake8
-    language_version: "python2"
-    exclude: '^src/acc/libsmm_acc/libcusmm/(tune|predict)_'
-  - id: flake8
-    language_version: "python3"
   - id: check-ast
     stages: [manual]
     language_version: "python2"
     exclude: '^src/acc/libsmm_acc/libcusmm/(tune|predict)_'
+  - id: flake8
+    language_version: "python3"
   - id: check-ast
     stages: [manual]
     language_version: "python3"

.pre-commit/check_header.py

@@ -17,9 +17,9 @@
 from contextlib import contextmanager
 
 TYPES = {
-    'c_cpp': ['.c', 'h', '.cc', '.hh', '.cxx', '.hpp', '.cu'],
-    'python': ['.py'],
-    'fortran': ['.F', '.f', '.f90', '.f03'],
+    "c_cpp": [".c", "h", ".cc", ".hh", ".cxx", ".hpp", ".cu"],
+    "python": [".py"],
+    "fortran": [".F", ".f", ".f90", ".f03"],
 }
 
 # max number of lines allowed between header and top of file
@@ -30,9 +30,9 @@
 
 
 @contextmanager
-def mmap_open(name, mode='r'):
-    access = mmap.ACCESS_READ if mode == 'r' else mmap.ACCESS_WRITE
-    with open(name, mode + 'b') as fhandle:
+def mmap_open(name, mode="r"):
+    access = mmap.ACCESS_READ if mode == "r" else mmap.ACCESS_WRITE
+    with open(name, mode + "b") as fhandle:
         fmapped = mmap.mmap(fhandle.fileno(), 0, access=access)
         yield fmapped
         fmapped.close()
@@ -44,7 +44,7 @@ def check_header(header_dir, files, verbose=False):
     header_len = {}
 
     for headertype in TYPES:
-        with open(path.join(header_dir, headertype), 'rb') as fhandle:
+        with open(path.join(header_dir, headertype), "rb") as fhandle:
             header_content = fhandle.read()
             header_re[headertype] = re.compile(re.escape(header_content))
             header_len[headertype] = len(header_content)
@@ -61,16 +61,22 @@ def check_header(header_dir, files, verbose=False):
 
         with mmap_open(fpath) as fmapped:
             header_type = ext_map[fext]
-            match = header_re[header_type].search(fmapped, 0, ALLOWED_LINES * MAX_LINE_LENGTH + header_len[header_type])
+            match = header_re[header_type].search(
+                fmapped, 0, ALLOWED_LINES * MAX_LINE_LENGTH + header_len[header_type]
+            )
 
             if not match:
                 print("✗ {} ... required header not found".format(fpath))
                 retval = 1
                 continue
 
-            lines_above = fmapped[0:match.start()].splitlines()
+            lines_above = fmapped[0 : match.start()].splitlines()
             if len(lines_above) > ALLOWED_LINES:
-                print("✗ {} ... header not within first {} lines".format(fpath, ALLOWED_LINES))
+                print(
+                    "✗ {} ... header not within first {} lines".format(
+                        fpath, ALLOWED_LINES
+                    )
+                )
                 retval = 1
                 continue
 
@@ -80,11 +86,13 @@ def check_header(header_dir, files, verbose=False):
     sys.exit(retval)
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     parser = argparse.ArgumentParser(description="Check files for header presence")
-    parser.add_argument('files', metavar='FILE', type=str, nargs='+', help="files to check")
-    parser.add_argument('--verbose', '-v', action='store_true', default=False)
+    parser.add_argument(
+        "files", metavar="FILE", type=str, nargs="+", help="files to check"
+    )
+    parser.add_argument("--verbose", "-v", action="store_true", default=False)
     args = parser.parse_args()
 
-    header_dir = path.join(path.dirname(path.abspath(__file__)), 'headers')
+    header_dir = path.join(path.dirname(path.abspath(__file__)), "headers")
     check_header(header_dir, args.files, args.verbose)

.travis.yml

@@ -90,15 +90,13 @@ jobs:  # the following jobs inherit only the first OS
     - make test
     - bash <(curl -s https://codecov.io/bash)
   - name: "Run pre-commit hook on changed files"
+    if: tag IS NOT present
     script:
-    - set -o pipefail
-      # fetch remotes first to ensure that referenced branches (other than the default) are available
-    - git config remote.origin.fetch "+refs/heads/*:refs/remotes/origin/*"
-    - git fetch
-    # yapf requires python3.6.4+
+    # ensure python 3.6.4+
     - pyenv global 3.6.7 2.7.15
-    - git diff --name-only --diff-filter=AM HEAD "origin/${TRAVIS_BRANCH}" | xargs pre-commit run --files
+    - pre-commit run --all-files
   - name: "Run pre-commit hook 'check-ast' on all files"
+    if: tag IS NOT present
     script:
     - pre-commit run --hook-stage manual --all-files check-ast
   - stage: release

CMakeLists.txt

@@ -13,6 +13,7 @@ cmake_dependent_option(WITH_C_API "Build the C API (ISO_C_BINDINGS)" ON "USE_MPI
 cmake_dependent_option(WITH_EXAMPLES "Build the examples" ON "USE_MPI" OFF) # all examples require MPI
 
 set(TEST_MPI_RANKS "auto" CACHE STRING "Number of MPI ranks for testing")
+set(TEST_OMP_THREADS 2 CACHE STRING "Number of OpenMP threads for testing")
 
 set(USE_SMM "blas" CACHE STRING "Small Matrix Multiplication implementation to use (default: blas)")
 set_property(CACHE USE_SMM PROPERTY STRINGS blas libxsmm)
@@ -96,7 +97,7 @@ if (Fortran_COMPILER_NAME MATCHES "ftn")
 endif ()
 
 # Depending on the fortran compiler, set the appropriate compiler flags
-if (Fortran_COMPILER_NAME MATCHES "gfortran.*")
+if (Fortran_COMPILER_NAME MATCHES "^gfortran.*")
   set(CMAKE_CXX_FLAGS              "-std=c++11")
   set(CMAKE_CXX_FLAGS_RELEASE      "-O3 -funroll-loops")
   set(CMAKE_CXX_FLAGS_COVERAGE     "-O0 -fprofile-arcs -ftest-coverage")
@@ -114,15 +115,15 @@ if (Fortran_COMPILER_NAME MATCHES "gfortran.*")
       SET(CMAKE_EXE_LINKER_FLAGS  "${CMAKE_EXE_LINKER_FLAGS} -lgcov --coverage")
     endif ()
   endif ()
-elseif (Fortran_COMPILER_NAME MATCHES "ifort.*")
+elseif (Fortran_COMPILER_NAME MATCHES "^ifort.*")
   set(CMAKE_CXX_FLAGS_RELEASE      "-O3")
   set(CMAKE_CXX_FLAGS_DEBUG        "-O0 -debug")
   set(CMAKE_Fortran_FLAGS          "-free -stand f03 -fpp")
   # Disable the line-length-extension warning #5268
   set(CMAKE_Fortran_FLAGS_RELEASE  "-O3 -diag-disable=5268")
   set(CMAKE_Fortran_FLAGS_DEBUG    "-O0 -debug")
   set(F2008_COMPILER_FLAGS         "-stand f08")
-elseif (Fortran_COMPILER_NAME MATCHES "pgfortran.*")
+elseif (Fortran_COMPILER_NAME MATCHES "^pgf.*")
   set(CMAKE_CXX_FLAGS_RELEASE      "-fast")
   set(CMAKE_CXX_FLAGS_DEBUG        "-g")
   set(CMAKE_Fortran_FLAGS          "-Mfreeform -Mextend -Mallocatable=03")
@@ -134,8 +135,8 @@ else ()
 Unknown compiler, trying with just '-O2'/'-O0 -g'.\n\
 You will most likely have to pass additonal options for free-form Fortran 2003/2008 for your compiler!\n\
 Please open an issue at https://github.com/cp2k/dbcsr/issues with the reported compiler name and the required flags.")
-  message("CMAKE_Fortran_COMPILER full path: " ${CMAKE_Fortran_COMPILER})
-  message("Fortran_COMPILER_NAME: " ${Fortran_COMPILER_NAME})
+  message("-- Fortran_COMPILER_NAME: " ${Fortran_COMPILER_NAME})
+  message("-- CMAKE_Fortran_COMPILER full path: " ${CMAKE_Fortran_COMPILER})
 
   set(CMAKE_Fortran_FLAGS_RELEASE  "-O2")
   set(CMAKE_Fortran_FLAGS_DEBUG    "-O0 -g")

CONTRIBUTING.md

@@ -1,11 +1,11 @@
 # Contributing to DBCSR
 The core of DBCSR is written in Fortran. All other languages must be supported through bindings. 
 
-There is a single [API](https://github.com/cp2k/dbcsr/blob/develop/src/dbcsr_api.F) file for DBCSR, which is provided for external usage only. **Do not use the API for any internal DBCSR development!** Packages build on top of DBCSR, for example [DBCSR Tensors](https://github.com/cp2k/dbcsr/tree/develop/src/tensors), **must only use** the DBCSR API.
+There is a single [API](./src/dbcsr_api.F) file for DBCSR, which is provided for external usage only. **Do not use the API for any internal DBCSR development!** Packages build on top of DBCSR, for example [DBCSR Tensors](./src/tensors), **must only use** the DBCSR API. Note that any change in the APIs will require a major release of the library.
 
-We support Make and CMake for compilation, please keep the build system updated when adding/removing files. When adding new functions, it is extremely important to provide simple test programs, aka "unit tests", to check whether these functions are performing as they should. The directory [test](https://github.com/cp2k/dbcsr/tree/develop/tests) serves as infrastructure for that. If you do not feel comfortable with integrating these tests with the build system, please notify the other developers.
+We support Make and CMake for compilation, please keep the build system updated when adding/removing files. When adding new functions, it is extremely important to provide simple test programs, aka "unit tests", to check whether these functions are performing as they should. The directory [test](./tests) serves as infrastructure for that. If you do not feel comfortable with integrating these tests with the build system, please notify the other developers.
 
-It is also appreciated to have examples (under the directory [examples](https://github.com/cp2k/dbcsr/tree/develop/examples)). They must be independent of the DBCSR compilation and only use the DBCSR API. 
+It is also appreciated to have examples (under the directory [examples](./examples)). They must be independent of the DBCSR compilation and only use the DBCSR APIs.
 
 DBCSR developers can find additional information on the [Development](https://github.com/cp2k/dbcsr/wiki/Development) wiki page.
 
@@ -19,10 +19,11 @@ Please make sure that you follow the following code conventions (based on [CP2K
 3. Every static variable should be marked with the `SAVE` attribute.
 4. Every Fortran module should contain the line `IMPLICIT NONE`.
 5. Every conversion that might change value should be explicit.
-6. Each `.F` file should contain either a `PROGRAM` or a single `MODULE`, whose name matches the filename. Then, it should start with the DBCSR header. Note that the name of the modules must be unique, even across different directories! 
-7. Use the routines from [MPI wrappers](https://github.com/cp2k/dbcsr/tree/develop/src/mpi) instead of calling MPI directly.
+6. Each `.F` file should contain either a `PROGRAM` or a single `MODULE`, whose name must start with the `dbcsr_` suffix and matches the filename. Then, it should start with the DBCSR header. Note that the name of the modules must be unique, even across different directories! 
+7. Use the routines from [MPI wrappers](./src/mpi) instead of calling MPI directly.
 8. Don't use `UNIT=*` in `WRITE` or `PRINT` statements. Instead request a unit from the logger:     `iw=dbcsr_logger_get_default_unit_nr()` and write only if you actually received a unit: `IF(iw>0) WRITE (UNIT=iw, ,,,)`.
 9. Avoid to use `STOP`. Prefer the DBCSR error handlers: `DBCSR_WARN`, `DBCSR_ABORT`, `DBCSR_ASSERT`. 
-10. Each preprocessor flag should start with two underscores and be documented in the [Makefile.inc](https://github.com/cp2k/dbcsr/blob/develop/Makefile.inc).
+10. Each preprocessor flag should start with two underscores and be documented in the [Makefile.inc](./Makefile.inc).
+11. All routines in the API must start with the `dbcsr_` namespace. For submodules API (e.g. [DBCSR Tensors](./src/tensors)), each function has to start with the `dbcsr_<unique ID of the submodule>_` namespace.
 
 **Most important, please avoid to commit dead code and useless comments!**

README.md

@@ -79,4 +79,4 @@ The configuration flags are (default first):
 ## Contributing to DBCSR
 
 Your contribution to the project is welcome! 
-Please see [Contributing.md](https://github.com/cp2k/dbcsr/blob/develop/CONTRIBUTING.md) and this [wiki page](https://github.com/cp2k/dbcsr/wiki/Development). For any help, please notify the other developers.
+Please see [Contributing.md](./CONTRIBUTING.md) and this [wiki page](https://github.com/cp2k/dbcsr/wiki/Development). For any help, please notify the other developers.

VERSION

@@ -1,6 +1,6 @@
-MAJOR = 1
-MINOR = 1
-PATCH = 0
+MAJOR = 2
+MINOR = 0
+PATCH = 0-alpha
 # A specific DATE (YYYY-MM-DD) fixes an official release, otherwise
 # it is considered Development version.
-DATE  = 2019-04-08
+DATE  = 2019-04-15